Analyzing the Regional Impact
of a Fossil Energy Cap in China
Da Zhang, Valerie Karplus, Sebastian Rausch and Xiliang Zhang
TSINGHUA - MIT
China Energy & Climate Project
Report No. 237
January 2013
The MIT Joint Program on the Science and Policy of Global Change is an organization for research, independent policy
analysis, and public education in global environmental change. It seeks to provide leadership in understanding scientific,
economic, and ecological aspects of this difficult issue, and combining them into policy assessments that serve the needs
of ongoing national and international discussions. To this end, the Program brings together an interdisciplinary group from
two established research centers at MIT: the Center for Global Change Science (CGCS) and the Center for Energy and
Environmental Policy Research (CEEPR). These two centers bridge many key areas of the needed intellectual work, and
additional essential areas are covered by other MIT departments, by collaboration with the Ecosystems Center of the Marine
Biology Laboratory (MBL) at Woods Hole, and by short- and long-term visitors to the Program. The Program involves
sponsorship and active participation by industry, government, and non-profit organizations.
To inform processes of policy development and implementation, climate change research needs to focus on improving the
prediction of those variables that are most relevant to economic, social, and environmental effects. In turn, the greenhouse
gas and atmospheric aerosol assumptions underlying climate analysis need to be related to the economic, technological, and
political forces that drive emissions, and to the results of international agreements and mitigation. Further, assessments of
possible societal and ecosystem impacts, and analysis of mitigation strategies, need to be based on realistic evaluation of the
uncertainties of climate science.
This report is one of a series intended to communicate research results and improve public understanding of climate issues,
thereby contributing to informed debate about the climate issue, the uncertainties, and the economic and social implications
of policy alternatives. Titles in the Report Series to date are listed on the inside back cover.
Ronald G. Prinn and John M. Reilly
Program Co-Directors
For more information, please contact the Joint Program Office
Postal Address: Joint Program on the Science and Policy of Global Change
77 Massachusetts Avenue
MIT E19-411
Cambridge MA 02139-4307 (USA)
Location:
400 Main Street, Cambridge
Building E19, Room 411
Massachusetts Institute of Technology
Access:
Phone: +1.617. 253.7492
Fax: +1.617.253.9845
E-mail: globalchange@mit.edu
Web site: http://globalchange.mit.edu/
Printed on recycled paper
Analyzing the Regional Impact of a Fossil Energy Cap in China
Da Zhang*†§ , Valerie Karplus* , Sebastian Rausch*‡§ , and Xiliang Zhang†
Abstract
Decoupling fossil energy demand from economic growth is crucial to China’s sustainable development. In
addition to energy and carbon intensity targets enacted under the Twelfth Five-Year Plan (2011–2015), a
coal or fossil energy cap is under discussion as a way to constrain the absolute quantity of energy used.
Importantly, implementation of such a cap may be compatible with existing policies and institutions. We
evaluate the efficiency and distributional implications of alternative energy cap designs using a numerical
general equilibrium model of China’s economy, built on the 2007 regional input-output tables for China and
the Global Trade Analysis Project global data set. We find that a national cap on fossil energy implemented
through a tax on final energy products and an energy saving allowance trading market is the most costeffective design, while a regional coal-only cap is the least cost-effective design. We further find that a
regional coal cap results in large welfare losses in some provinces. Capping fossil energy use at the national
level is found to be nearly as cost effective as a national CO2 emissions target that penalizes energy use
based on carbon content.
Contents
1. INTRODUCTION ................................................................................................................................... 1
2. CONTEXT FOR ANALYZING AN ENERGY CAP POLICY IN CHINA ................................. 3
2.1 Energy Cap Background ................................................................................................................ 3
2.2 Literature Review ............................................................................................................................ 4
3. HETEROGENEITY OF ENERGY USE ACROSS CHINA’S PROVINCES: A
MULTI-REGIONAL INPUT-OUTPUT ANALYSIS OF EMBODIED ENERGY ...................... 5
3.1 Data and Methodology ................................................................................................................... 5
3.2 MRIO Results .................................................................................................................................. 6
4. MODEL AND SCENARIOS ................................................................................................................ 7
4.1 Overview of Model Structure ....................................................................................................... 7
4.1.1 Firm and Household Behavior ......................................................................................... 10
4.1.2 Supplies of Final Goods and Treatment of Domestic and International Trade ........ 12
4.1.3 Equilibrium and Model Solution ..................................................................................... 13
4.2 Model Scenarios ............................................................................................................................ 13
5. RESULTS ............................................................................................................................................... 15
5.1 National Impacts ........................................................................................................................... 15
5.2 Regional Impacts .......................................................................................................................... 16
5.3 Policy Design Choices Under an Energy Cap .......................................................................... 17
5.3.1 National versus regional targets ...................................................................................... 18
5.3.2 Primary versus final energy use ....................................................................................... 18
5.4 Impact on Non-Fossil Energy ..................................................................................................... 19
6. CONCLUSIONS ................................................................................................................................... 20
7. REFERENCES ...................................................................................................................................... 23
1. INTRODUCTION
In the search for policy approaches to address environmental, climate, and health damages
from fossil energy use, China has been the first country to propose capping the use of coal and
*
Joint Program of the Science and Policy of Global Change, Massachusetts Institute of Technology, Cambridge, U.S.
Institute of Energy, Environment, and Economy, Tsinghua University, China.
§
Corresponding author (Email: zhangda@mit.edu).
‡
Department of Management, Technology and Economics, ETH Zurich, Zurich, Switzerland.
†
1
other fossil fuels directly.1 While the precise features of the cap have yet to be determined,
policymakers have suggested that the cap will restrict primary consumption of coal to 3.9 billion
tons of coal equivalent in 2015, compared to 3.5 billion tons used in 2011 (Kraemer, 2012;
National Development and Reform Commission, 2012).2 It is not clear whether or not this cap
will be difficult to meet, as it depends on the rates of China’s economic growth and energy
intensity improvement. If the high growth rates (greater than 10%) of recent years persist, keeping
the rate of increase in coal use to less than 3% per year will require a significant reduction in the
intensity of coal use. Given the potentially central role of an energy cap in China’s future climate
policy, this analysis investigates the cost effectiveness and distributional impact of a coal or fossil
energy cap as an energy use and carbon dioxide (CO2 ) reduction strategy.
Recent policy discussion has centered on whether a cap should restrict the use of coal only or
restrict all fossil energy sources. Advocates of a coal cap argue that coal is the appropriate policy
target because the environmental and health externalities of coal use are the most detrimental, and
because there is a history of setting specific goals for coal capacity and use. Constraining all fossil
sources, by contrast, would allow energy reductions to come from the natural gas, oil, and coal
sectors. Since fuels are not differentiated based on carbon content under an energy cap, there is
concern that capping all fossil sources might perversely result in a situation where relatively
cleaner natural gas is reduced at the expense of continued reliance on coal. By simulating both
types of caps in a numerical energy-economic general equilibrium model and comparing them to
a national cap on CO2 embodied in energy, we investigate the existence and magnitude of such
effects, and to compare the relative efficiency of different policy designs.
China’s economy and energy system are characterized by significant heterogeneity—coal
production is concentrated in the northern and western regions, while demand for electricity and
energy for industrial processes is highest along the more prosperous eastern seaboard. The
availability of low carbon substitutes for fossil energy such as hydropower and wind are also
regionally determined. The country’s western provinces, in contrast to those in the east, are less
affluent and, while they are major suppliers of energy, they use relatively less total energy and use
energy less efficiently, in part due to reliance on inefficient production technology and in part due
to their economic structures (Liu et al., 2012). Policymakers broadly agree that climate policy
should not exaggerate these regional disparities, and if possible should accelerate sustainable
development and technological upgrading in lagging regions. In this analysis we are able to
investigate how the impact of a coal or energy cap would be distributed, and if alternative policy
designs could help to improve overall cost effectiveness and minimize undesirable distributional
impacts.
1
A cap could take the form of a constraint on coal use or other forms of fossil energy. Although often simply referred
to as an energy cap, the policy under consideration exempts energy types with no associated carbon footprint, such
as nuclear and renewable energy. The cap would limit on an energy basis the total fossil energy use, which in
China is typically measured in metric tons of coal equivalent (Xinhua Press, 2011).
2
The Twelfth Five-Year Plan’s Coal Industry Development Plan assigns a target for China’s coal production capacity
of 4.1 billion tons, but a target for coal consumption of 3.9 billion tons in 2015.
2
This paper is organized as follows. Section 2 provides background and context for analyzing
an energy cap policy in China, reviewing the current policy discussion and important policy
design considerations. Section 3 discusses regional energy and economic system characteristics
across China’s provinces, providing a basis for interpreting the comparison of policy scenarios.
Section 4 describes the model used in this analysis. Section 5 presents the results of the analysis,
focusing on the cost effectiveness and distributional impacts of alternative policy designs. Section
6 discusses the implications of the results for policy design decisions, which in practice will
inevitably weight cost effectiveness against the political feasibility of various options.
2. CONTEXT FOR ANALYZING AN ENERGY CAP POLICY IN CHINA
2.1 Energy Cap Background
Most of the discussion around limiting energy use in China has focused on coal (Point Carbon,
2012). Coal supplied around 77% of Chinas primary energy demand in 2010 (National Statistics
Bureau, 2011b). Coal also has the highest carbon content of all fossil fuels and has been
responsible for localized air and water pollution as well as severe health consequences across
China (Matus et al., 2012). The notion of a coal cap is actually not new, and dates back to
historical practice of production targets chosen within the context of China’s planned economy. In
China’s recent five-year plans, the stated coal consumption target has been consistently exceeded
but is still considered as an important policy instrument (Zhang et al., 2012b). In the Twelfth
Five-Year Plan (2011 to 2015), a target for coal use at the national level has been set, restricting
primary consumption to 3.9 billion tons in 2015. This cap would be implemented through targets
set at the provincial (or even sub-provincial levels), as it is established practice in China’s
administrative system to explicitly assign responsibility for meeting national targets to individual
provinces. Several provinces and municipalities have already announced a cap on coal for the
Twelfth Five-Year Plan (People’s Daily Online, 2011; 21st Century Business Herald, 2011;
Xinhua Press, 2011).
Recently, policymakers have discussed replacing or complementing the existing coal cap with
a fossil energy cap, with the objectives of curbing energy use, reducing CO2 emissions, and
improving air quality and health. While a target would be assigned at the national level (Lan,
2012; Reuters, 2012), this proposal is also likely to include the allocation of targets at the
provincial level. There are no reports that any allowance trading mechanism has been discussed in
relation to a coal or fossil energy cap, although in policy discourse a fossil energy cap has been
discussed as a step toward the development of a national cap-and-trade system.
Another design question is whether the cap will restrict primary energy or final energy in order
to achieve a targeted reduction in total energy use—in all existing energy cap proposals to date,
primary energy is the target. Primary energy may be easier to target than final energy in the
current institutional setup, given that oversight involves controlling fewer entities, many of which
have direct links to national or provincial governments. Targeting final energy demand would
allow reductions to be achieved by limiting both local primary energy and imported energy
embodied in final demand. Economic intuition would suggest that broader coverage—i.e.
3
including all fossil energy sources, allowing trading under a national cap, and targeting reductions
in final rather than primary energy demand—would be the most cost-effective approach, given
that it maximizes abatement flexibility. However, implementing these provisions will require
some changes in China’s institutional environment that could prove difficult to implement on a
short time frame or lack the support necessary to be implemented effectively. In this context, it is
important to understand both the efficiency penalties associated with moving to less flexible
policy designs as well as the distributional implications.
An investigation of the cost effectiveness and distributional impacts of various energy cap
policy designs requires a model that captures in a single framework sufficient detail in China’s
energy and economic system at the regional level. It must also capture interactions across multiple
markets and endogenous changes in prices and income, given that an energy cap would bear on
one or more energy commodities that are inputs to a large number of sectors across the economy.
We therefore employ a regional computable general equilibrium model of China that represents
economic flows as well as energy demands and associated CO2 emissions as supplemental
physical accounts. This model allows us to assess the full range of policy design combinations
and compare their effects on energy use, CO2 emissions, and the economy at the national and
provincial levels. Our outputs are intended to supply a cost-effectiveness comparison that can be
considered alongside political feasibility and other criteria in the policy design process.
2.2 Literature Review
Our analysis of the coal cap contributes to a rapidly growing literature on the design of energy
and climate policy in China (Zhang, 1998, 2000; Cao et al., 2009; Li et al., 2009; Cong and Wei,
2010; He et al., 2010; Lu et al., 2010; Zhao and Ortolano, 2010; Dai et al., 2011; Zhang et al.,
2011; Zhou et al., 2011; Wei et al., 2012; Zhang et al., 2012a). Methodologies used to evaluate
policies vary widely, from system dynamics modeling (Cong and Wei, 2010) to CGE modeling
(Zhang, 1998; He et al., 2010; Dai et al., 2011; Zhang et al., 2012a) to qualitative investigation
(Zhao and Ortolano, 2010; Zhang et al., 2011). Many studies use observations of regional energy
intensities and abatement costs to inform the design of environmental policy in China (Zhou
et al., 2011; Wei et al., 2012). Other scholarship has focused on how differences in abatement
potential and costs across provinces can be used to determine how reduction targets are assigned.
Wei et al. (2012) provides estimates of abatement potential and costs for 29 provinces, with the
goal of balancing efficiency and equity concerns through the assignment of targets and in the
absence of an inter-provincial trading mechanism.
Given the importance of understanding the consequences of policy design choices for regional
energy use, emissions, and economic activity, we focus on the role of three salient policy design
choices currently under discussion—whether to cap coal only or all fossil energy sources,
whether to assign targets based on primary or final energy consumption, and whether to assign
targets at the national or regional level. Specifically, we are interested in understanding the
consequences of targeting primary or final energy use to achieve the intended reductions, an issue
discussed at length in (Bushnell, 2011). We are also concerned with the consequences of
4
assigning regional (e.g., provincial) or national targets, which was considered for the case of
China in a study of CO2 intensity target design (Zhang et al., 2012a). Our investigation
contributes to the literature on energy and climate policy design in China by simulating a full
range of alternative energy cap designs in a comprehensive energy-economic framework. Given
that addressing climate change is an important policy objective, we further compare simulated
energy cap policy designs to a cap on CO2 emissions.
3. HETEROGENEITY OF ENERGY USE ACROSS CHINA’S PROVINCES: A MULTIREGIONAL INPUT-OUTPUT ANALYSIS OF EMBODIED ENERGY
3.1 Data and Methodology
To develop intuition to assist in interpreting the results of the policy analysis, we undertake an
empirical assessment of embodied energy use across China’s provinces. This analysis is based on
an energy-economic data set that includes a consistent representation of energy markets in
physical units as well as detailed accounts of regional production and bilateral trade for the year
2007. The data set is based on detailed provincial-level data for China and a global economic and
energy data set. Data for China is based on the full set of China’s recently published 2007
provincial input-output (IO) tables for all 30 provinces3 and China’s national IO table (National
Statistics Bureau, 2011a). The IO tables for each province differentiate 42 sectors, and include
data on various existing tax rates in the Chinese economy. The global data is based on the GTAP
database (GTAP, 2012), version 8, which identifies 129 countries and regions and 57
commodities, providing consistent global accounts of production, consumption and bilateral trade
as well as consistent accounts of physical energy flows, energy prices and emissions in the year
2007. Energy use and emissions data are based on data from GTAP (GTAP, 2012) and the 2007
China Energy Statistical Yearbook (National Statistics Bureau, 2008).
Compared to the 42 sectors in the China’s provincial IO tables, there are only 29 sectors
represented in the energy balance tables. To get the largest resolution, we aggregate IO tables
using all the overlapping sectors with energy balance tables to 29 sectors.4 The energy balance
tables for each province of China contain only aggregated energy use data for all 24 secondary
industry sectors.
The China Energy Statistical Yearbook also provides detailed energy use data for each
secondary industry at the national level. We use these two sources to estimate energy use data for
each secondary industry at the provincial level. All the energy data is measured by more detailed
types of energy products. Using standard conversion factors for China (National Statistics
Bureau, 2008), we aggregate the energy data to six energy product types (Coal, Refined oil and
coal products, Crude oil, Natural gas, Fuel gas, Electricity and heat) which are consistent with the
energy sectors in provincial IO tables.5
3
4
5
Tibet is not included because of lack of data and its small economic size.
The sector mapping is available on request from the authors.
The precise mapping of energy products in the energy balance tables to energy products in IO tables is available by
request from the authors.
5
Integration of different data sources is accomplished using least-square optimization
techniques, and is described in detail in Zhang et al. (2012a). We find substantial discrepancies
between the provincial- and national-level economic accounts, consistent with the findings by
Guan et al. (2012). Our estimation routine benchmarks the aggregate of provincial-level accounts
to national totals, as national accounting methods are more standardized than provincial
accounting methods. We then use the provincial data to describe the shares of economic activity
in each province. To integrate physical energy flows from official statistics within the regional
SAM accounts, we replace energy-market related entries by imputed energy value (money) flows;
these are calculated as the product of physical flows and energy price data supplied by the Energy
Research Institute of the National Development and Reform Commission (NDRC).
Following the methodology detailed in Böhringer et al. (2011), we calculate embodied energy
use for each primary energy type, accounting for indirect energy use associated with non-fossil
inputs in addition to direct energy use. For this calculation we define the following multi-region
IO model (MRIO). Let produced goods, comprising fossil fuels indexed by f , be indexed by i,
y
and Ef,g,r
denote embodied energy f for the joint set of activities—produced goods, final demand,
m
describes embodied energy
investment, and government demand denoted by g in province r. Ei,r
of imported commodities defined as a weighted average of imported varieties across trade
partners. In our integrated economy-energy data set, the following accounting identity holds:
X
X
ag,r =
bi,g,r +
ci,g,r + dg,r
(1)
i
i
where a, b, and c denote the values of output, domestic intermediate inputs, and imported
intermediate inputs, respectively. d is the sum of factor and tax payments with which no energy
use is associated. The multi-regional IO model relates the E variables to equation 1. Embodied
energy use in output is thus given by:
X y
X
y
m
Ef,g,r
ag,r = Direct Energyf,g,r +
Ef,i,r bi,g,r +
Ef,i,r
ci,g,r .
(2)
i
|
{zi
Indirect Energy
}
Equation 2 can be represented as a square system of equations which we solve recursively using a
diagonalization algorithm (Böhringer et al., 2011).
3.2 MRIO Results
Previous studies have analyzed the heterogeneity in patterns of energy use and CO2 emissions
across China’s provinces (Wang, 2011; Guo et al., 2012; Liu et al., 2012). Differences in CO2
intensity across provinces have been related to economic structure and the relative efficiency of
production technology (Liu et al., 2012), although a broad range of factors are important. This
section investigates the heterogeneity of China’s regional energy system by considering two
relevant metrics, specifically the coal use intensity across China’s provinces and energy intensity
of final demand. This data analysis provides a foundation for interpreting model results described
6
later in the paper.
Table 1 summarizes the energy and electricity embodied in total output by province.
Embodied energy by type differs substantially across provinces, and reflects the prevailing
technologies and structure of the economy in each region. The intensity of coal use in provincial
output varies substantially due to differences in economic structure and the intensity of coal
use—the highly developed regions of Guangdong, Shanghai, Beijing, and Jiangsu have the lowest
intensity of coal use and have diversified, developed economies, while regions where coal is
abundant and technology is less developed (Qinghai, Shanxi, Guizhou, Neimenggu, and Ningxia)
have a relative high coal use intensity.
Provinces also differ greatly in both the composition and the efficiency of electric power
generation. Some provinces have significant room to improve the fossil fuel efficiency of
generation through fuel switching, investment in efficiency upgrades, or new construction of low
carbon renewable or nuclear generation. The availability and relative cost of these opportunities
will determine to a large extent the cost of abatement in the electricity sector of each province.
Figure 1 shows the intensity of fossil energy use embodied in the unit production of the
electricity and heat sector in each province. Hubei, Ningxia and Sichuan have large shares of
hydroelectric generation that do not show up in the fossil energy total, leading them to appear less
energy intensive overall.
Provinces also differ significantly in terms of the total amount of energy embodied in final
demand by households, as well as in its intensity, as shown in Figure 2. Final energy refers to the
energy embodied in both internally produced goods and net imports that are consumed by
household within each province. As Figure 2 shows, the provinces that consume large amounts
of energy are among the least energy intensive. For example, Guangdong has the highest total
household energy use but the lowest energy intensity, and Shandong, Zhejiang, and Liaoning
exhibit similar patterns. By contrast, Ningxia, Hainan, and Qinghai have among the highest
energy intensities of household consumption but the lowest total energy use by households. Thus
a policy that disproportionately targets populous provinces with a high level of energy
consumption may miss significant (and likely cost-effective) energy saving opportunities.
4. MODEL AND SCENARIOS
We employ the China Regional Energy Model (C-REM), a multi-commodity, multi-regional
static numerical general equilibrium model of the world economy with sub-national detail for
China’s economy (Zhang et al., 2012a). C-REM was jointly developed through a collaboration
between the Institute for Energy, Environment, and Economy at Tsinghua University and the MIT
Joint Program on the Science and Policy of Global Change. For this study, we aggregate the data
set to 30 provinces in China and six world regions (see Table 2), and into 26 commodity groups
distinguishing six energy goods (coal, natural gas, fuel gas, crude oil, refined oil and electricity)
and 13 non-energy commodities (see Table 3). The key features of the model are outlined below.
7
Table 1. Intensity of primary fuel or electricity use embodied in output by province compared to China’s
average and the world average levels. Units are million tons of coal equivalents per billion USD in 2007.
Province
Location
Coal
Crude oil
Natural Gas
Electricity
Guangdong
Shanghai
Beijing
Jiangsu
Hainan
Fujian
Zhejiang
Shandong
Guangxi
Sichuan
Hunan
Hubei
Tianjin
Henan
Liaoning
Shaanxi
Jiangxi
Chongqing
Anhui
Heilongjiang
Jilin
Xinjiang
Hebei
Yunnan
Gansu
Ningxia
Neimenggu
Guizhou
Shanxi
Qinghai
South
East
North
East
South
East
East
East
South
Southwest
Central
Central
North
Central
Northeast
Northwest
East
Southwest
East
Northeast
Northeast
Northwest
North
Southwest
Northwest
Northwest
North
Southwest
North
Northwest
0.32
0.33
0.36
0.38
0.40
0.41
0.48
0.52
0.52
0.53
0.57
0.57
0.58
0.61
0.61
0.63
0.64
0.65
0.66
0.68
0.75
0.75
0.76
0.78
0.80
0.84
0.98
1.13
1.35
1.67
0.19
0.28
0.19
0.18
0.50
0.17
0.19
0.21
0.16
0.13
0.19
0.22
0.31
0.12
0.35
0.30
0.19
0.14
0.17
0.29
0.29
0.42
0.20
0.18
0.45
0.24
0.17
0.16
0.15
0.22
0.04
0.06
0.05
0.03
0.18
0.02
0.03
0.02
0.01
0.06
0.02
0.02
0.05
0.03
0.03
0.04
0.02
0.09
0.03
0.05
0.03
0.13
0.03
0.03
0.13
0.09
0.04
0.04
0.03
0.09
0.10
0.09
0.10
0.10
0.13
0.10
0.13
0.12
0.11
0.11
0.11
0.13
0.12
0.11
0.12
0.11
0.12
0.12
0.11
0.10
0.13
0.18
0.13
0.12
0.19
0.25
0.14
0.16
0.20
0.28
0.56
0.12
0.21
0.15
0.04
0.06
0.12
0.05
China–average
World–average
8
Ningxia
Hainan
Qinghai
Gansu
Guangxi
Yunnan
Chongqing
Guizhou
Tianjin
Jiangxi
Neimenggu
Xinjiang
Shaanxi
Beijing
Shanxi
Fujian
Anhui
Sichuan
Jiangsu
Shanghai
Hunan
Hubei
Heilongjiang
Henan
Hebei
Liaoning
Jilin
Zhejiang
Shandong
Guangdong
Energy emobied in consumption (Mtce)
70
Gas
Crude oil
Coal
50
40
1.5
30
1
20
0
9
Energy intensity (mtce/billion USD)
Hubei
Ningxia
Sichuan
Guangxi
Gansu
Guangdong
Beijing
Hunan
Fujian
Shanghai
Chongqing
Zhejiang
Hainan
Jiangsu
Yunnan
Shaanxi
Tianjin
Guizhou
Xinjiang
Liaoning
Shandong
Shanxi
Hebei
Qinghai
Jiangxi
Neimenggu
Henan
Anhui
Heilongjiang
Jilin
Energy use per unit of electricity production (Mtce/Mtce)
3.5
3
2.5
2
1.5
1
0.5
0
Figure 1. Primary fuel use embodied in the per unit production of electricity and heat, by province
(mtce/mtce).
2.5
60
Eint(mtce/billion $)
2
10
0.5
0
Figure 2. Total and composition of household final energy consumption by fuel type and energy intensity,
by province.
Table 2. Regional aggregation in the model.
China’s provinces
Other regions
30 provinces∗
EUR: Europe (EU 27 + EFTA)
APD: Asian and Pacific developed countries or regions∗∗
ASI: Other Asian countries
NAM: North America (U.S. and Canada)
CAA: Central Asia, Africa and Former Soviet Union
ROW: Latin America and rest of the world
* including 4 municipal cities and 5 autonomous regions, and Tibet is not included due to data availability.
** including Japan, Korea, Singapore, Taiwan, Australia and New Zealand.
4.1 Overview of Model Structure
4.1.1 Firm and Household Behavior
For each industry (i = 1, . . . , I, i = j) in each region (r = 1, . . . , R) gross output (Yir ) is
produced using inputs of labor (Lir ), capital (Kir ), natural resources including coal, natural gas,
crude oil, and land (Rir ), and produced intermediate inputs (Xjir )6 :
Yir = Fir (Lir , Kir , Rir ; X1ir , . . . , XIir ) .
(3)
We employ constant-elasticity-of-substitution (CES) functions to characterize the production
technologies. All industries are characterized by constant returns to scale and are traded in
perfectly competitive markets. Nesting structures for the production and consumption systems are
detailed in Zhang et al. (2012a).
Fossil fuels f (coal, crude oil and natural gas) are produced according to a nested CES
function combining a fuel-specific resource, capital, labor, and intermediate inputs:
Yf r =
ρR
αf r Rf frr
ρR
fr
1/ρRfr
+ νf r min (X1f r , . . . , Xif r , Vf r )
(4)
where α, ν are share coefficients of the CES function and σfRr = 1/(1 − ρR
f r ) is the elasticity of
substitution between the fuel-specific resource and the composite including primary factors,
energy and materials. σfRr is determined by the resource input share and price elasticity of supply
ηf r . The primary factor and energy composite is a Cobb-Douglas function of the energy input,
labor and capital:
Vf r = Lβf 1r Kfβr2 E1βe1
. . . Eiβei
(5)
fr
fr
where β1 , β2 , βe1 , . . . , βei are shares of the labor, capital and energy inputs. Oil refining, gas
production and distribution production are represented in Figure 3.
6
For simplicity, we abstract from the various tax rates that are used in the model. The model includes ad valorem
output taxes and import tariffs.
10
Gross output i
σ=0
KLE
σkle
Energy
σenoe
Non-ELE
σen
M1
CRU
···
Mj
···
MJ
Capital-Labor
σva
ELE
K
L
COL GAS GDT
(a)
Gross output i
σ=0
KLE
σkle
Energy
σenoe
Non-ELE
σen
OIL
COL
CRU
GAS
M1
···
Mj
···
Capital-Labor
σva
ELE K
L
GDT
(b)
Figure 3. Structure of production for oil refining i ∈ {OIL} (a) and gas production and distribution
i ∈ {GDT} (b).
11
MJ
Table 3. Sectoral aggregation in the model.
Sector
Abbreviation
Sectors aggregated from GTAP 8 data set
Agriculture
AGR
Coal
Crude oil
Natural gas
Minerals mining
Light industries
Petroleum, coal products
Energy intensive industries
Transport equipment
Other manufacturing industries
Electricity
Gas manufacture and distribution
Water
Trade
Transport
Other service industry
COL
CRU
GAS
OMN
LID
OIL
EID
TME
OID
ELE
GDT
WTR
TRD
TRP
OTH
PDR, WHT, GRO, V F, OSD, C B, PFB,
OCR, CTL, OAP, RMK, WOL, FRS, FSH,
CMT, OMT, VOL, MIL, PCR, SGR
COL
CRU
GAS
OMN
FBT, TEX, WAP, LEA, LUM, PPP
OIL
CRP, NMM, I S, NFM
MVH, OTN
FMP, OME, ELE, OMF, CNS
ELY
GDT
WTR
TRD
OTP, WTP, ATP
CMN, OFI, ISR, OBS, ROS, OSG, DWE
We distinguish several generation technologies, including conventional fossil, hydro, nuclear
and wind. In this version of the model, the resource input share is calibrated using the benchmark
data. As we lack estimates of price elasticities for supply of nuclear, hydro, and wind in
individual provinces in China, we adopt the corresponding elasticities from the MIT Emissions
Prediction and Policy Analysis model (Paltsev et al., 2005).
For each sector, the capital mobility feature is represented by following a putty-clay approach.
A fraction φ of previously-installed capital becomes non-malleable in each sector, and vintaged
production in this sector uses this part of capital with fixed shares of all the inputs which are
identical to those installed in the base year. The fraction 1 − φ of capital is malleable and can be
shifted to other sectors in response to input price changes. All the sectors except electricity have
the same φ value, while φ for the electricity sector is higher because capital tends to be less
mobile when invested in electricity generation (Sue Wing, 2006).
In each region r, preferences of representative consumers are represented by a Leontief utility
function comprised of consumption goods (Ci ) and investment (I):
Ur = min [g(C1r , . . . , CIr ), g(I1r , . . . , IIr )]
where the function g(·) is a CES composite of all goods. In each region, a single government
entity approximates government activities at both central and local levels.
12
(6)
4.1.2 Supplies of Final Goods and Treatment of Domestic and International Trade
All intermediate and final consumption goods are differentiated following the Armington
assumption. For each demand class, the total supply of good i is a CES composite of a
domestically produced variety and an imported variety, as follows:
i1/ρDi
ρD
i
ZMir
(7)
i D
h
ρD
ρD 1/ρi
Cir = ψ c CDiri + ξ c CMiri
(8)
h
Xir = ψ
h
z
Iir = ψ
i
ρD
ZDiri
ρD
IDiri
+ξ
+ξ
z
i
i1/ρDi
ρD
i
IMir
i1/ρDi
h
ρD
ρD
g
g
i
i
Gir = ψ GDir + ξ GMir
(9)
(10)
where Z, C, I and G are inter-industry demand, consumer demand, investment demand, and
government demand for good i, respectively; and ZD, ZM, CD, CM, ID, IM, GD, GM are
domestic and imported components of each demand class, respectively. The ψ’s and ξ’s are the
CES share coefficients. The Armington substitution elasticities between domestic and imported
varieties in these composites are given by σiD = 1/(1 − ρD
i ).
The domestic and imported varieties of goods are represented by nested CES functions. We
replicate a border effect within our Armington import specification by assuming that goods
produced within China are closer substitutes than goods from international sources. We include
separate import specifications for China’s provinces (indexed by p = 1, . . . , P ) and international
regions (indexed by t = 1, . . . , T ). More detail about the nesting structure of the Armington
composites is provided in Zhang et al. (2012a).
4.1.3 Equilibrium and Model Solution
Consumption, labor supply and savings result from the decisions of the representative
household in each model region that maximize its utility subject to a budget constraint that
consumption equals income. Given input prices gross of taxes, firms maximize profits subject to
the technology constraints. Firms are assumed to operate in perfectly competitive markets (an
assumption that can be relaxed in specific applications) and maximize profit by selling products at
a price equal to the marginal cost of production. Numerically, the equilibrium is formulated as a
mixed complementarity problem (MCP) (Mathiesen, 1985; Rutherford, 1995). A model solution
must satisfy zero profit and market clearance conditions, with the former condition determining a
vector of activity levels and the latter a vector of market-clearing prices. The problem is
formulated in GAMS and solved using the mathematical programming system MPSGE
(Rutherford, 1999) and the PATH solver (Dirkse and Ferris, 1995) to obtain non-negative prices
and quantities.
13
4.2 Model Scenarios
Given that we are interested in comparing the impacts of policies on a consistent basis, we
model all energy cap policies such that they achieve a 20% reduction in China’s fossil energy use
embodied in final demand of all sectors, households, government and investment activities.7 In
reality the cap is likely to be set several years in advance and allow a fixed increase in coal use
relative to a reference year (for instance, the 2015 coal cap allows an increase in coal energy use
of 0.4 billion tons relative to 2011).8 Given our comparative-static model, we choose a 20%
reduction as a representative reduction target to compare the impact of alternative designs.
To represent the energy cap in the C-REM model, an endogenous tax is levied on coal or fossil
energy use whose level is determined in equilibrium by the constraint on absolute energy
consumption. For the national target aimed at limiting fossil energy use at the national level, a
nation-wide energy saving allowance trading market is established which allows some provinces
to reduce more energy use and sell unused allowances to other provinces. The initial allowance
allocation is based on the total final energy consumption for each province.
The objective of a first set of scenarios is to quantify the national and regional changes in the
energy mix as well as the associated welfare impacts of constraining energy use through a coal
cap relative to a fossil energy cap. These scenarios are defined as follows:
• COAL CAP: Achieves a 20% reduction in final fossil energy use at the national level by
capping coal in primary energy use and implementing the cap through provincial targets
(with no coal saving allowance trading market).
• FOSSIL CAP: Achieves a 20% reduction in final fossil energy use at the national level by
capping all final fossil energy use and allowing energy saving allowance trading across
provinces.
• CO2 CAP: Achieves the same national reduction in carbon emissions as FOSSIL CAP by
capping carbon emissions and allowing emissions allowance trading across provinces.
The coal cap as modeled here (targeted at primary energy at the provincial level, no
inter-provincial trading) is based on current designs under discussion. For an initial comparison
with a fossil energy cap, we target a similar energy reduction at the national level through a cap
on all forms of fossil energy use (coal, natural gas, and oil) and include an energy saving
allowance trading market, before later exploring alternative policy designs that allow less
abatement flexibility. Given that the objective of the energy cap is in part to address carbon
emissions, we then compare the fossil energy cap to a cap on carbon emissions that achieves a
carbon reduction identical to that achieved under the fossil energy cap. By targeting carbon
7
8
A final demand constraint is not directly applied to energy used in intermediate energy production, e.g., electricity
generation and coke production.
The coal industry policy for the Twelfth Five-Year Plan suggests that the target is meant to be approximate, with
“more or less” (zuoyou in Chinese) after the 3.9 billion ton target.
14
emissions directly the carbon policy would be expected to achieve emissions reductions at least
cost by discriminating among energy sources on the basis of carbon content.
To allow for a comparison among alternative fossil energy cap designs, we consider variants of
a fossil energy cap with more limited abatement flexibility—either by assigning targets at the
regional level, targeting reductions through primary energy only, or both. We therefore consider
an additional set of scenarios:
• NCAP FINALDEMAND: Same as FOSSIL CAP above, achieves a 20% reduction in
final fossil energy use at the national level by capping all final fossil energy use and
allowing energy saving allowance trading across provinces.
• NCAP PRIMDEMAND: Achieves a 20% reduction in final fossil energy use at the
national level by capping all primary fossil energy use and allowing energy saving
allowance trading across provinces.
• RCAP FINALDEMAND: Achieves a 20% reduction in final fossil energy use at the
provincial level by capping all final fossil energy use with no trading across provinces.
• RCAP PRIMDEMAND: Achieves a 20% reduction in final fossil energy use at the
provincial level by capping all primary fossil energy use with no trading across provinces.
5. RESULTS
To investigate the national and regional impacts of the coal and fossil energy cap policies, we
consider the changes in the energy mix and total final fossil energy use as well as the regional
distribution of impacts. Coal and fossil energy cap designs are first compared to each other, and
then the fossil energy cap is compared to a national cap on energy-embodied CO2 emissions. We
then consider alternative designs for the fossil energy cap, such as a provincial-level
implementation (without an allowance market) and imposing the target on primary rather than
final energy use. The alternative design choices may improve the ease of policy implementation.
We then quantify any associated cost penalty.
5.1 National Impacts
Comparing the impact of a coal cap and a fossil energy cap imposed at the national level, we
find that a coal cap that reduces fossil energy use by 20% has the direct effect of reducing coal use
by 31%. A coal cap also indirectly causes a reduction of other fossil energy types because higher
coal prices increase the cost of final goods, reducing overall demand. Under a coal cap, demand
for refined oil and coke products and natural gas falls by about 5%. A fossil energy cap also
reduces coal use, but by only 26%, while natural gas demand falls by 19% and refined oil and
coke products by around 8%. The modest reduction in oil use reflects its high marginal cost of
abatement, given that there are few substitutes for oil in the parts of the economy where it is
primarily used, such as the transportation sector.
15
We compare the results of the fossil energy cap policy to a constraint on carbon embodied in
energy use that achieves an identical reduction in emissions, and find that reductions differ only
slightly. Coal is reduced by 27%, while gas falls by 14% and refined oil and coke products by 7%.
The difference in the contribution of each energy type reflects the fact that coal has a higher
carbon content than oil or natural gas, and thus under a policy that targets carbon rather than
energy, it is cost effective to further reduce coal. However, it is interesting that the fossil energy
cap—a more blunt instrument—produces an outcome that is close to the carbon policy. Our
model results suggest that, of all the energy sources, coal can be reduced most cost effectively,
whether it is on a carbon or an energy basis.
Turning to the welfare impacts, we find that, at the national level, the coal cap causes the
largest loss in welfare, measured as equivalent variation (EV) relative to reference, at (3.2%),
while both the fossil energy cap and the carbon policy produce a loss in EV of only 1.3%. The
largest welfare loss occurs under the coal cap because it requires that all reductions come from
coal use reduction and ignores opportunities to cost effectively reduce energy by constraining
other fossil sources. This same logic holds true for carbon reduction, although it should be noted
that a coal cap achieves a larger carbon reduction than an equivalent fossil energy cap because of
coal’s proportionally higher carbon content. Nevertheless, even under a CO2 emissions cap, coal
is not the only contributor to carbon reduction—cost-effective opportunities to reduce emissions
by curbing use of oil and natural gas also form part of the solution.
5.2 Regional Impacts
We now turn to examine policy impacts at the regional level. The distributional impacts of
each policy will reflect the geography of use for each energy source as well as diversity in the
marginal abatement cost. Figure 4 shows final fossil energy use under each of the policy
scenarios compared to a no policy reference case. We find that, relative to either a fossil energy
cap or carbon policy, the coal cap results in significantly larger reductions in coal use in some
relatively affluent provinces (Beijing, Shanghai, Guangdong, Hebei,and Jiangsu). By contrast,
Yunnan, Xinjiang, Inner Mongolia, Guizhou and Gansu—all resource-rich, less developed
provinces—would reduce coal more under a national fossil energy or carbon target, reflecting an
abundance of low cost opportunities in these provinces.
We now consider how changes in energy use across provinces are reflected in welfare changes.
Figure 5 shows the consumption loss under the coal cap, fossil energy cap and carbon cap
policies by province. It is perhaps unsurprising that under the coal cap we find large concentrated
welfare costs in provinces that face high marginal costs of abatement, either because they have
already achieved significant efficiency in using coal (Beijing, Shanghai, Guangdong, Jiangsu and
Shandong) or because they have few competitive substitutes for the coal they already use (Beijing
and Shandong) or because they are main coal producers (Shanxi and Shaanxi). For instance,
Guangdong and Jiangsu have highly developed and relatively modern industrial and electricity
sectors that have already realized significant coal use efficiency. In these cases, coal cap will have
significant negative impacts on economy and household welfare.
16
180
Oil
Processed Energy Products
Natural Gas
Coal
160
140
Energy use (mtce)
120
100
80
60
40
20
Yunnan
Zhejiang
Tianjin
Xinjiang
Shanxi
Sichuan
Shanghai
Shaanxi
Shandong
Ningxia
Qinghai
Neimenggu
Jiangxi
Liaoning
Jilin
Jiangsu
Hubei
Hunan
Heilongjiang
Hebei
Henan
Hainan
Guangxi
Guizhou
Guangdong
Fujian
Gansu
Chongqing
Anhui
Beijing
0
Figure 4. Final fossil energy use under each of the policy scenarios compared to a no policy reference
case (from left to right for each province: no policy, COAL CAP, FOSSIL CAP and CO2 CAP).
Both the fossil energy cap and the carbon cap result in similar patterns of welfare loss across
provinces. This result is not surprising given that the only difference between the two policies is
that the carbon cap penalizes embodied carbon rather than energy use alone, and the overall
contribution of the various energy types is similar. In many provinces, losses under a fossil energy
or carbon cap are less pronounced relative to the losses under the coal cap, as both the fossil
energy cap and the carbon cap are designed to allow inter-provincial trading that takes advantage
of the large, concentrated availability of abatement opportunities in some provinces. As
mentioned above, Gansu, Guizhou, Qinghai and Yunnan all play an important role in reductions
under a fossil energy cap or carbon cap (see Figure 4), and thus the welfare effects in these cases
are slightly more pronounced (see Figure 5). As the coal cap does not fully take advantage of low
cost abatement opportunities in these provinces, they are actually better off under a coal
cap—despite the fact that, at the national level, welfare loss is significantly larger. The
consumption losses in the coal-rich provinces Shanxi and Shaanxi are the most insensitive to the
choice of policy, reflecting the fact that the provincial coal cap target is similar to the
cost-effective provincial contribution that results under a national fossil energy cap.
5.3 Policy Design Choices Under an Energy Cap
We now consider how changing the implementation of a fossil energy cap affects both the
energy mix composition and welfare outcomes at the national and regional levels. Specifically,
applying targets at the regional level may be politically attractive in China, given that economic
and other goals (environmental, health, public safety) are evaluated at the provincial level.
17
10
Consumption Change (%)
5
0
-5
-10
-15
COAL_CAP
FOSSIL_CAP
CO2_CAP
Beijing
Shanxi
Shandong
Neimenggu
Chongqing
Sichuan
Shaanxi
Hainan
Shanghai
Hebei
Heilongjiang
Tianjin
Hunan
Guangdong
Zhejiang
Ningxia
Xinjiang
Henan
Jiangxi
Anhui
Jilin
Liaoning
Jiangsu
Guangxi
Gansu
Fujian
Hubei
Qinghai
Yunnan
Guizhou
-20
Figure 5. Consumption change in percent under the COAL CAP, FOSSIL CAP and CO2 CAP scenarios.
Developing a trading mechanism would require the development of new institutions to allow the
verification and payment for abatement undertaken outside of provincial borders. It may also
prove more feasible to cap primary, rather than final, energy consumption, given that primary
energy use is easier to track. The purpose of this exercise is to quantify the welfare impact of
moving to alternative policy configurations in order to make explicit the tradeoffs inherent in
choices facing policymakers.
5.3.1 National versus regional targets
Regional targets spread the burden of reducing fossil energy use equally across provinces, but
with the result that abatement is undertaken at different marginal costs, raising the overall cost of
the policy. At the national level, the welfare penalty for moving from a national to a regional
(provincial) fossil energy cap is a 26% increase in the welfare loss to −1.6%. Provinces with
relatively high marginal abatement costs contribute more than would be expected under regional
targets. Figure 6 shows how this welfare impact is distributed across provinces. Most provinces
incur higher welfare costs (particularly in Beijing, Tianjin, Shanxi, Shandong and Xinjiang). In
general moving from national to regional targets does not alter the direction of the welfare effects
by province. The corresponding impact on the primary energy mix is shown in Figure 7.
5.3.2 Primary versus final energy use
Capping final energy instead of primary energy has more pronounced effects on welfare,
particularly when combined with regional targets. At the national level, targeting primary energy
at the national level increases the welfare loss by 34% relative to a national policy that targets
18
10
Consumption change (%)
5
0
-5
-10
RCAP_PRIMDEMAND
NCAP_PRIMDEMAND
-15
RCAP_FINALDEMAND
NCAP_FINALDEMAND
Shanxi
Heilongjiang
Shaanxi
Xinjiang
Chongqing
Beijing
Neimenggu
Tianjin
Liaoning
Hainan
Hunan
Ningxia
Sichuan
Zhejiang
Anhui
Jiangxi
Shandong
Jilin
Henan
Hebei
Guangdong
Gansu
Jiangsu
Guangxi
Shanghai
Fujian
Hubei
Qinghai
Guizhou
Yunnan
-20
Figure 6. Relatuve consumption change in percent under the three variants of the Fossil Energy Cap
policy compared to NCAP FINALDEMAND.
final demand; if the policy is implemented at the regional level, the incremental welfare loss
increases by 84%. The regional distribution of impacts changes dramatically as well (Figure 6). A
regional policy that caps primary fossil energy changes the direction of welfare effects in Shanxi,
Liaoning and Yunnan, in some cases by very large magnitudes. These differences likely reflect
the distinct geographies associated with primary fossil energy use and fossil energy used in final
demand. Where abatement opportunities and costs within a single province differ widely between
primary and final energy demand, the choice of target can have large effects on welfare.9
5.4 Impact on Non-Fossil Energy
As mentioned previously an energy cap would not be applied on non-fossil energy sources,
including electricity generation from nuclear, hydro and wind resources. However an energy cap
could affect the use of these sources if these non-fossil sources prove to be cost-effective
substitutes for coal or fossil energy types. Indeed, we find that all policies considered result in an
increase in non-fossil energy use in every province. Shanghai, Shandong and Jiangsu in particular
see strong increases in non-fossil energy use under the regional or coal caps, relative to the
9
We further performed a sensitivity analysis to consider the impact of changing target stringency on the relative
cost effectiveness of the policy designs considered. Compared to a base case reduction of 20%, increasing the
target stringency to 30% or reducing it to 10% did not affect the cost-effectiveness ranking of policy designs.
However, we do note that, as target stringency increases, targeting primary rather than final energy leads to a larger
incremental welfare decrease, especially in the regional policy scenarios.
19
180
Oil
Processed Energy Products
Natural Gas
Coal
160
140
Energy use (mtce)
120
100
80
60
40
20
Yunnan
Zhejiang
Tianjin
Xinjiang
Shanxi
Sichuan
Shanghai
Shaanxi
Shandong
Ningxia
Qinghai
Neimenggu
Jiangxi
Liaoning
Jilin
Jiangsu
Hubei
Hunan
Heilongjiang
Hebei
Henan
Hainan
Guangxi
Guizhou
Guangdong
Fujian
Gansu
Chongqing
Anhui
Beijing
0
Figure 7. Final fossil energy use under the four variants of the Fossil Energy Cap policy (from left to right
for each province: NCAP FINALDEMAND, RCAP FINALDEMAND, NCAP PRIMDEMAND and
RCAP PRIMDEMAND).
national policies or carbon cap.10
6. CONCLUSIONS
If the objective of policy is to reduce fossil energy use, we find that a coal cap as currently
proposed (e.g., on primary energy demand at the regional level and without energy-saving
allowance trading) is less cost effective than a fossil energy cap. This conclusion is robust to a
comparison with all variants of a fossil energy cap modeled in this analysis.11 A fossil energy cap
that achieves an equivalent reduction in energy is more cost effective than a coal cap because it
allows cost-effective reductions to be achieved through abatement of other carbon-intensive
energy sources. If the objective is to reduce CO2 emissions, we find that a national fossil energy
cap that targets energy in final demand is almost as efficient as a CO2 emissions cap. This is
because coal is a major source of low cost reductions in both energy and emissions, and
contributes significantly (but not exclusively) to reductions under both policies. Excluding
10
One caveat is that non-fossil electricity generation technology is not represented as backstop technologies in the
model. Because of the constant elasticity of substitution (CES) function representation of technology, this implies
that if there is no nuclear, hydro and wind electricity generation in the reference scenario, i.e. an initial share of
zero in the CES function, there will be no new installation in the policy scenarios. The representation of advanced
non-fossil electricity generation options in the model is beyond the scope of this paper.
11
For all cases, we assume that the costs of compliance with policy are passed through to final users. However, China
has managed electricity pricing (Lam, 2004), and so some end use decisions would not respond to the policy,
reducing overall cost effectiveness.
20
carbon-intensive primary fuels, such as natural gas and oil, from an energy cap indirectly
subsidizes these fuels, which could in turn offset the climate benefits associated with capping coal
alone.
This finding may not be intuitive to those who point out that a fossil energy cap would
indirectly penalize low carbon fuels, such as natural gas, which it is argued would be phased out
first due to their high costs of production. Our model results emphasize that it is essential not to
conflate cost of production with marginal abatement cost, the latter of which determines where
reductions in fuel use can be undertaken most cost effectively. Indeed, we find that natural gas is
reduced only slightly more under a fossil energy cap relative to a carbon cap, and the opposite is
found for coal, which is reduced slightly less. As coal is inexpensive, the shadow price on energy
(or carbon) causes a large increase in the price of coal in percentage terms; natural gas, which is
already relatively costly, increases in price by a smaller relative percentage. Regardless of whether
energy or CO2 is the target, reducing coal use proves to be a dominant abatement strategy. The
distributional impacts of both policies are also qualitatively and quantitatively similar.
Regardless of cost effectiveness, in practice policies that build on existing institutions and
practices may stand a greater chance of implementation. A coal cap may be easier to implement
because it focuses on a single commodity, and has perhaps the strongest policy precedent in
China. However, as our analysis indicates, a similar amount of energy can be reduced through a
national fossil energy cap at only 40% of the cost of a coal-only cap.12 When it comes to the
design of a fossil energy cap, a national constraint on fossil energy embodied in final demand
(with an energy-saving allowance trading system) is the most cost effective. Assigning targets at
the regional level and/or on primary energy use carries an incremental welfare penalty of 27% to
84%, relative to a policy that assigns a national constraint that targets final energy demand.
Despite these welfare penalties, alternative policy designs may prove more politically feasible.
Policy targets in China have traditionally been implemented through regional channels, with
enforcement delegated to provincial governments. We find that moving from national to regional
fossil energy targets carries a modest welfare penalty and does not substantially change the
distribution of impacts relative to the national target. The welfare penalty is larger when moving
to a target on primary energy, particularly if it is implemented at the regional level. The welfare
penalties associated with alternative policy designs should be explicitly considered in the policy
process.
While a fossil energy cap may be a reasonable substitute for a direct constraint on carbon in
the target range we consider here, it would not address another major concern of China’s
policymakers—energy security threats associated with reliance on imported oil. Neither a fossil
energy cap nor a carbon cap would lead to significant reduction in petroleum demand, given the
lack of substitutes for its use in transportation and chemical production. If energy security is the
12
The magnitude of this advantage depends on the design of the coal and fossil energy cap. We also model a national
cap on coal (not shown) and find that it results in a lower welfare loss of 2.7%. This loss is still greater than the
loss expected under the least cost effective of the fossil energy cap designs modeled, a regional cap that acts by
constraining primary fossil energy, which has a welfare loss of −2.3%.
21
desired objective, complementary policies will be needed—for instance, a tax or cap on
petroleum use as well as incentives to develop alternatives to petroleum-intensive activities. This
analysis underscores that a sharp policy instrument is always preferred to a blunt instrument, but
that some blunt instruments are sharper than others.
Acknowledgements
We acknowledge the support of the Ministry of Science and Technology of China through the
Institute for Energy, Environment, and Economy at Tsinghua University, and the support of the
Graduate School at Tsinghua University, which are supporting Zhang Da’s doctoral research as a
visiting scholar at the Massachusetts Institute of Technology. We further thank Eni S.p.A., ICF
International, Shell International Limited, and the French Development Agency (AFD), founding
sponsors of the China Energy and Climate Project. We also grateful for support provided by the
Social Science Key Research Program from National Social Science Foundation, China of Grant
No. 09&ZD029 and by Rio Tinto China. We would further like to thank John Reilly, Sergey
Paltsev, Kyung-min Nam, Henry Chen, Paul Kishimoto and Audrey Resutek for helpful
comments, discussion and edits.
22
7. REFERENCES
21st Century Business Herald, 2011: Nine regions have set initial coal consumption caps.
Böhringer, C., J. C. Carbone and T. F. Rutherford, 2011: Embodied Carbon Tariffs.
Bushnell, E. T., James B.and Mansur, 2011: Vertical targeting and leakage in carbon policy. The
American Economic Review, 101(3): 263–267.
Cao, J., M. S. Ho and D. W. Jorgenson, 2009: The local and global benefits of green tax policies
in China. Review of Environmental Economics and Policy, 3: 189–208.
Cong, R.-G. and Y.-M. Wei, 2010: Potential impact of (CET) carbon emissions trading on
China’s power sector: A perspective from different allowance allocation options. Energy,
35(9): 3921–3931.
Dai, H., T. Masui, Y. Matsuoka and S. Fujimori, 2011: Assessment of China’s climate
commitment and non-fossil energy plan towards 2020 using hybrid AIM/CGE model. Energy
Policy, 39(5): 2875–2887.
Dirkse, S. P. and M. C. Ferris, 1995: The PATH Solver: a non-monontone stabilization scheme
for Mixed Complementarity Problems. Optimization Methods and Software, 5: 123–156.
GTAP, 2012: Global Trade, Assistance, and Production: The GTAP 8 data base. Center for
Global Trade Analysis, Purdue University.
Guan, D., Z. Liu, Y. Geng, S. Lindner and K. Hubacek, 2012: The gigatonne gap in China’s
carbon dioxide inventories. Nature Climate Change, 2: 672–674.
Guo, J., Z. Zhang and L. Meng, 2012: China’s provincial CO2 emissions embodied in
international and interprovincial trade. Energy Policy, 42: 486–497.
He, Y. X., S. L. Zhang, L. Y. Yang, Y. J. Wang and J. Wang, 2010: Economic analysis of coal
price - electricity price adjustment in China based on the CGE model. Energy Policy, 38(11):
6629–6637.
Kraemer, S., 2012: China to Simply Cap Coal Use Within 3 Years. Scientific American, March 9.
Lam, P.-L., 2004: Pricing of electricity in China. Energy, 29: 287–300.
Lan, L., 2012: Energy use may be capped for 2015. People’s Daily Online.
Li, N., M.-j. Shi and F. Wang, 2009: Roles of regional differences and linkages on Chinese
regional policy effect in CGE analysis. Systems Engineering - Theory and Practice, 29:
35–44.
Liu, Z., Y. Geng, S. Lindner and D. Guan, 2012: Uncovering China’s greenhouse gas emission
from regional and sectoral perspectives. Energy, 45(1): 1059–1068.
23
Lu, C., Q. Tong and X. Liu, 2010: The impacts of carbon tax and complementary policies on
Chinese economy. Energy Policy, 38: 7278–7285.
Mathiesen, L., 1985: Computation of economic equilibria by a sequence of linear
complementarity problems. Mathematical Programming Study, 23: 144–162.
Matus, K., K.-M. Na, N. E. Selin, L. N. Lamsal, J. M. Reilly and S. Paltsev, 2012: Health
damages from air pollution in China. Global Environmental Change, 22: 5566.
National Statistics Bureau, 2008: 2007 China Energy Statistical Yearbook.
National Statistics Bureau, 2011a: 2007 China Regional Input-Output Tables.
National Statistics Bureau, 2011b: 2010 China Statistical Yearbook.
NDRC, 2012: Twelfth Five-Year Plan Coal Industry Development Plan (In Chinese).
Paltsev, S., J. Reilly, H. Jacoby, R. Eckaus, J. McFarland, M. Sarofim, M. Asadoorian and
M. Babiker, 2005: The MIT Emissions Prediction and Policy Analysis (EPPA) Model:
Version 4. MIT JPSPGC Report 125, August, 72 p.
(http://mit.edu/globalchange/www/MITJPSPGC Rpt125.pdf ).
People’s Daily Online, 2011: Wu Yin: China’s coal consumption in 2015 should be controlled
within 3.9 billion ton.
Point Carbon, 2012: China’s Tianjin sets 2015 coal consumption cap. Point Carbon - Reuters.
Reuters, 2012: China may miss new target to cut coal output growth.
Rutherford, T. F., 1995: Extension of GAMS for complementarity problems arising in applied
economics. Journal of Economic Dynamics and Control, 19(8): 1299–1324.
Rutherford, T. F., 1999: Applied general equilibrium modeling with MPSGE as a GAMS
subsystem: an overview of the modeling framework and syntax. Computational Economics,
14: 1–46.
Sue Wing, 2006: The synthesis of bottom-up and top-down approaches to climate policy
modeling: Electric power technologies and the cost of limiting US CO2 emissions. Energy
Policy, 34: 3847–3869.
Wang, X., 2011: On China’s energy intensity statistics: Toward a comprehensive and transparent
indicator. Energy Policy, 39(11): 7284–7289.
Wei, C., J. Ni and L. Du, 2012: Regional allocation of carbon dioxide abatement in China. China
Economic Review, 23(3): 552–565.
24
Xinhua Press, 2011: Shandong Province will cap coal consumption during the Twelfth Five-Year
Plan.
Zhang, D., S. Rausch, V. Karplus and X. Zhang, 2012a: Quantifying regional economic impacts
of CO2 intensity targets in China. MIT JPSPGC Report 230, September, 26 p.
(http://globalchange.mit.edu/files/document/MITJPSPG Rpt230.pdf ).
Zhang, F., Y. Guo and X. Chen, 2011: Research on China’s power sector carbon emissions
trading mechanism. Energy Procedia, 12(0): 127–132.
Zhang, Q., K. He and H. Huo, 2012b: Cleaning China’s air. Nature, 484: 161–162.
Zhang, Z., 1998: Macroeconomic effects of CO2 emission limits: A computable general
equilibrium analysis for China. Journal of Policy Modeling, 20(2): 213–250.
Zhang, Z., 2000: Can China afford to commit itself an emissions cap? An economic and political
analysis. Energy Economics, 22: 587–614.
Zhao, X. and L. Ortolano, 2010: Implementing China’s national energy conservation policies at
state-owned electric power generation plants. Energy Policy, 38(10): 6293–6306.
Zhou, J., M. Duan and C. Liu, 2011: Output-based allowance allocations under China’s carbon
intensity target. Energy Procedia, 5(0): 1904–1909.
25
REPORT SERIES of the MIT Joint Program on the Science and Policy of Global Change
FOR THE COMPLETE LIST OF JOINT PROGRAM REPORTS:
http://globalchange.mit.edu/pubs/all-reports.php
192. The Impact of Border Carbon Adjustments under
Alternative Producer Responses Winchester February
2011
193. What to Expect from Sectoral Trading: A U.S.-China
Example Gavard et al. February 2011
194. General Equilibrium, Electricity Generation
Technologies and the Cost of Carbon Abatement Lanz
and Rausch February 2011
195. A Method for Calculating Reference
Evapotranspiration on Daily Time Scales Farmer et al.
February 2011
196. Health Damages from Air Pollution in China
Matus et al. March 2011
197. The Prospects for Coal-to-Liquid Conversion:
A General Equilibrium Analysis Chen et al. May 2011
198. The Impact of Climate Policy on U.S. Aviation
Winchester et al. May 2011
199. Future Yield Growth: What Evidence from Historical
Data Gitiaux et al. May 2011
200. A Strategy for a Global Observing System for
Verification of National Greenhouse Gas Emissions
Prinn et al. June 2011
201. Russia’s Natural Gas Export Potential up to 2050
Paltsev July 2011
202. Distributional Impacts of Carbon Pricing: A General
Equilibrium Approach with Micro-Data for Households
Rausch et al. July 2011
203. Global Aerosol Health Impacts: Quantifying
Uncertainties Selin et al. August 201
204. Implementation of a Cloud Radiative Adjustment
Method to Change the Climate Sensitivity of CAM3
Sokolov and Monier September 2011
205. Quantifying the Likelihood of Regional Climate
Change: A Hybridized Approach Schlosser et al. October
2011
206. Process Modeling of Global Soil Nitrous Oxide
Emissions Saikawa et al. October 2011
207. The Influence of Shale Gas on U.S. Energy and
Environmental Policy Jacoby et al. November 2011
208. Influence of Air Quality Model Resolution on
Uncertainty Associated with Health Impacts Thompson
and Selin December 2011
209. Characterization of Wind Power Resource in the
United States and its Intermittency Gunturu and
Schlosser December 2011
210. Potential Direct and Indirect Effects of Global
Cellulosic Biofuel Production on Greenhouse Gas
Fluxes from Future Land-use Change Kicklighter et al.
March 2012
211. Emissions Pricing to Stabilize Global Climate Bosetti et
al. March 2012
212. Effects of Nitrogen Limitation on Hydrological
Processes in CLM4-CN Lee & Felzer March 2012
213. City-Size Distribution as a Function of Socioeconomic Conditions: An Eclectic Approach to Downscaling Global Population Nam & Reilly March 2012
214. CliCrop: a Crop Water-Stress and Irrigation Demand
Model for an Integrated Global Assessment Modeling
Approach Fant et al. April 2012
215. The Role of China in Mitigating Climate Change Paltsev
et al. April 2012
216. Applying Engineering and Fleet Detail to Represent
Passenger Vehicle Transport in a Computable General
Equilibrium Model Karplus et al. April 2012
217. Combining a New Vehicle Fuel Economy Standard with
a Cap-and-Trade Policy: Energy and Economic Impact in
the United States Karplus et al. April 2012
218. Permafrost, Lakes, and Climate-Warming Methane
Feedback: What is the Worst We Can Expect? Gao et al. May
2012
219. Valuing Climate Impacts in Integrated Assessment
Models: The MIT IGSM Reilly et al. May 2012
220. Leakage from Sub-national Climate Initiatives: The Case
of California Caron et al. May 2012
221. Green Growth and the Efficient Use of Natural Resources
Reilly June 2012
222. Modeling Water Withdrawal and Consumption for
Electricity Generation in the United States Strzepek et al.
June 2012
223. An Integrated Assessment Framework for Uncertainty
Studies in Global and Regional Climate Change: The MIT
IGSM Monier et al. June 2012
224. Cap-and-Trade Climate Policies with Price-Regulated
Industries: How Costly are Free Allowances? Lanz and
Rausch July 2012.
225. Distributional and Efficiency Impacts of Clean and
Renewable Energy Standards for Electricity Rausch and
Mowers July 2012.
226. The Economic, Energy, and GHG Emissions Impacts of
Proposed 2017–2025 Vehicle Fuel Economy Standards in
the United States Karplus and Paltsev July 2012
227. Impacts of Land-Use and Biofuels Policy on Climate:
Temperature and Localized Impacts Hallgren et al. August
2012
228. Carbon Tax Revenue and the Budget Deficit: A Win-WinWin Solution? Sebastian Rausch and John Reilly August 2012
229. CLM-AG: An Agriculture Module for the Community
Land Model version 3.5 Gueneau et al. September 2012
230. Quantifying Regional Economic Impacts of CO2
Intensity Targets in China Zhang et al. September 2012
231. The Future Energy and GHG Emissions Impact of
Alternative Personal Transportation Pathways in China
Kishimoto et al. September 2012
232. Will Economic Restructuring in China Reduce TradeEmbodied CO2 Emissions? Qi et al. October 2012
233. Climate Co-benefits of Tighter SO2 and NOx Regulations
in China Nam et al. October 2012
234. Shale Gas Production: Potential versus Actual GHG
Emissions O’Sullivan and Paltsev November 2012
235. Non-Nuclear, Low-Carbon, or Both? The Case of Taiwan
Chen December 2012
236. Modeling Water Resource Systems under Climate
Change: IGSM-WRS Strzepek et al. December 2012
237. Analyzing the Regional Impact of a Fossil Energy Cap in
China Zhang et al. January 2013
Contact the Joint Program Office to request a copy. The Report Series is distributed at no charge.